Three dimensional unassisted sit-to-stand prediction for virtual healthy young and elderly individuals

Sit-to-stand (STS) motion is one of the most important tasks in daily life and is one of the key determinants of functional independence, especially for the senior people. The STS motion has been extensively studied in the literature, mostly through experiments. Compared to numerous experimental studies, there are limited simulations with mostly assuming bilateral symmetry for STS tasks. However, it is not true even for healthy individuals to perform STS tasks with a perfect bilateral symmetry. In this study, predictive dynamics is utilized for STS prediction. The problem can be constructed as a nonlinear optimization formulation. The digital human model has 21 degrees of freedom (DOFs) for the unassisted STS tasks. The quartic B-spline interpolation is implemented for representing joint angle profiles. The recursive Lagrangian dynamics approach and the Denavit–Hartenberg method are implemented for the equations of motion. This study is to develop a generic three-dimensional unassisted STS motion prediction method for healthy young and elderly individuals. Results show that trunk joint angle peak values are similar between the two virtual-groups in the sagittal, frontal, and transverse planes. Lower-limbs’ joint angle and velocity profiles and their peak values between the right and left side for both virtual groups are also similar. The normalized peak joint torques are slight differences in each active DOF between the two virtual groups and the peak values are similar. The proposed method has been indirectly validated through the literature experimental results. The developed method has various potential applications in the design of exoskeleton, microelectromechanical system for fall detection, and assistive devices in rehabilitation.